Artificial Intelligence and Predictive Analytics in Diabetes Care

Artificial intelligence and machine learning are reshaping diabetes prevention, diagnosis, and management across the care continuum. Continuous glucose monitors, smart insulin pens, activity trackers, and dietary logging applications generate unprecedented volumes of longitudinal data; electronic health records contain rich clinical phenotypes; and multi-omics approaches offer new avenues for risk stratification. These data streams create opportunities for predictive models that identify individuals at risk for diabetes onset, forecast glycemic excursions, and anticipate complications before clinical manifestation. Yet persistent challenges remain regarding generalizability, transparency, clinical utility, and ensuring these tools benefit diverse populations.This Research Topic aims to highlight significant advances and outstanding challenges in harnessing AI and predictive analytics to improve outcomes for individuals with or at risk for type 1 diabetes, type 2 diabetes, and gestational diabetes. We seek to showcase innovations in risk prediction, glycemic pattern recognition, complication forecasting, and AI-driven clinical decision support, while emphasizing rigorous validation, health equity, and responsible implementation. Submissions may address the following themes:• Risk prediction and early detection: Machine learning models for identifying individuals at elevated risk for type 1 diabetes, type 2 diabetes, gestational diabetes, or progression from prediabetes; integration of genetic, metabolic, behavioral, and digital phenotyping data for personalized risk assessment• Glycemic forecasting and CGM analytics: Algorithms for predicting hypoglycemia and hyperglycemia; pattern recognition in continuous glucose data; time-in-range optimization; decision support for insulin dosing and lifestyle modifications• Complication prediction and prevention: Models predicting diabetic retinopathy, nephropathy, neuropathy, and cardiovascular events; computer vision applications for retinal screening and wound assessment• Clinical decision support and patient-facing tools: AI-driven tools embedded in clinical workflows; natural language processing for extracting diabetes-relevant information from clinical notes; large language model applications in patient education, clinical documentation, and self-management support• Methodological advances: Federated learning and privacy-preserving approaches; foundation models and domain adaptation; synthetic data generation; explainability and interpretability methods• Implementation and validation: Real-world deployment of predictive tools; external validation across diverse populations; addressing algorithmic bias and ensuring equitable performance; regulatory pathways and ethical considerationsWe encourage submissions that address health equity implications and strategies for responsible, equitable implementationConflict of Interest Declaration for Topic Editor David Liebovitz- Advisor to AI startups: Dendritic Health, KyralHealth, Optima Health for BP Topic editors: David Liebovitz, Yi Guo, Amisha Wallia Source: https://www.frontiersin.org/research-topics/76953/artificial-intelligence-and-predictive-analytics-in-diabetes-care